The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external UI files (e.g., CSS, JS, HTML) to identify raw styling values that should be replaced with tokens.
Ingestion points: scripts/audit_tokens.py reads file content from paths provided at runtime using the Path.read_text() method.
Capability inventory: The skill executes the scripts/audit_tokens.py script, which has the capability to read any file on the filesystem that the execution environment permits and print its contents to standard output.
Boundary markers: The skill does not implement delimiters or specific instructions to ignore malicious prompt content that might be embedded in the audited source code.
Sanitization: No sanitization is performed on the file content before it is processed by regex or printed to the console, although the script only targets specific line-based patterns.
[COMMAND_EXECUTION]: The skill instructions provide a command for the agent to run a bundled Python script (scripts/audit_tokens.py). This script is used for the primary purpose of auditing the project for design system compliance and does not interact with external networks or sensitive system configurations.

semantic-design-tokens